Effect of an elemental diet on body composition

GASTROENTEROLOGY

77:652-657.1979

Effect of an Elemental Diet on Body Composition A Comparison C. K. YEUNG, University

Department

with Intravenous

R. C. SMITH,

and

G.

Nutrition

L. HILL

of Surgery, The General Infirmary at Leeds, Leeds, England

Measurements of changes in body fat, protein, and water were carried out on two comparable groups of 14 ill surgical patients each over a 2-wk period during which one group received an elemental diet (nonprotein energy source was 67% carbohydrate and 33% fat) and the other a course of intravenous nutrition (nonprotein energy source was 200% carbohydrate). The patients fed with the elemental diet had no significant changes in body weight, fat protein, water, or plasma proteins over the study period, and although the patients fed intravenously also had no changes in body protein or plasma proteins, there was an average gain of 3.2 kg of body weight. This weight gain was mainly extracellular water. It is concluded that the administration of the elemental diet by continuous infusion was comparable to intravenous nutrition in maintaining body protein in these very ill patients and had the advantage of being cheaper and easier to manage. The problem of extracellular water accumulation seen in the patients fed intravenously was not present in the patients who received the elemental diet.

Intravenous nutrition has been shown to be capable of producing positive nitrogen balance and significant weight gain in critically ill patients.’ Weight gain, however, cannot always be equated with a gain in lean body tissue because the weight gain seen in patients fed intravenously with an amino acid and Received September 16,1976.Accepted April 24,1979. Address requests for reprints to: Mr. G. L. Hill, Ch. M., F.R.A.C.S., F.R.C.S., Senior Lecturer in Surgery, University Department of Surgery, The General Infirmary, Leeds 1,England. We would like to thank Dr. L. Burkinshaw, Dr. C. B. Oxby and the Department of Medical Physics, Dr. A. H. Smith (Department of Nuclear Medicine), Dr. I. McCarthy, Mr. A. Hackett, Professor D. Johnston (Department of Surgery), the Pharmacy Staff, and the Nursing Staff on Wards 27 and 28 of the Leeds General Infirmary for their help and encouragement. 0 1979 by the American Gastroenterological Association 0016-5065/79/100652-Os$OZ.oO

dextrose solution can be due to a gain in body wafer.z.3 An excessive gain in body water is undesirable as it may be an important contributory factor in the respiratory distress syndrome which is sometimes seen in critically ill patients on intravenous nutrition. The administration of an elemental diet by continuous intragastric or intrajejunal infusion can be an alternative to intravenous nutrition.4-7 It is therefore of importance to know whether the problem of retention of excessive body water also occurs with this form of therapy. In this study, we looked at the tissue composition of the weight changes in patients receiving a course of (a) an elemental diet and (b) intravenous nutrition.

Patients and Methods Measurement carried each

of changes

out on two groups over

a 2-wk

an elemental

diet

period (ED)

in body

of 14 ill adult during

and

the

which other

composition surgical

was

patients

one group

received

a course

of intra-

venous nutrition (IVN). The patients studied were a consecutive series who were presented to the University Department of Surgery for a course of nutritional support which lasted for approximately 2 wk. There was no attempt made to randomize or match the two groups of patients as such an attempt would have been very difficult in practice, and possibly unethical in some cases. The decision as to which of the two types of nutritional therapy was given to any individual patient was made purely on clinical grounds by one of the authors (G. L. Hill). The indications for nutritional support were varied in the 2 groups of patients and are shown in Table 1. There was one patient in each group who received nutritional support preoperatively because of marked weight loss (15% and 30% of the preillness weight, respectively). Four patients in the ED group and five patients in the IVN group were given prophylactic nutritional support after major abdominal surgery, and in these patients the feeding was initiated within 2-3 days after surgery. Two patients with a large bowel fistula were treated with the elemental

ELEMENTAL

Octohcr 1979

diet and both healed spontaneously without further surgery. One patient with a high output small bowel fistula was treated with intravenous nutrition, and, though the fistula failed to close, the clinical state of the patient was markedly improved to enable direct surgical closure. Three patients in the ED group and two in the IVN group had the short gut syndrome (
Body

Composition

BODY COMI’OSITION

653

rectly by the isotope dilution technique using 100 &i of tritiated water given as an intravenous injection. A blood sample was taken at 3 hr. and duplicate aliquots of the plasma were added to NE260 (Nuclear Enterprises Ltd., Edinburgh, Scotland) scintillator and measured on a liquid scintillation counter. For the second measurement of total body water, a predose background blood sample was taken for the measurement of remaining radioactivity. The standard deviation of repeated measurements for a subject of constant body composition was: -+ 0.4 kg for body weight, +- 0.65 kg for fat, f 0.36 kg for protein, f 48 mmol for potassium, and less than + 1.0 liter for total body water.

Plasma

Proteins

Plasma albumin, prealbumin, and transferrin levels were measured before and at the end of the 2 wk of feeding in both groups of patients. Plasma albumin was estimated by B.C.G. binding with a Vickers M300 analyzer,” and plasma prealbumin and transferrin were determined by radial immunodiffusion,‘” using plates prepared from commercially available antisera (Hoechst Pharmaceuticals, Middlesex, England). These investigations were approved by the Research Ethics Committee of the General Infirmary at Leeds (March 1976).

Measurements

Informed consent was obtained from all patients. Body composition was measured immediately before feeding was started and after 2 wk of feeding using the technique described in detail elsewhere.” In brief, the patients were weighed, and the thicknesses of the mid biceps, mid triceps, and subscapular skinfolds were measured three times each with a Holtain skinfold caliper. The mean values of the three measurements at these three sites were used to calculate body density using the method described by Durnin and Wormersley.” Body fat was subsequently derived using Siri’s equation.q Total body nitrogen was measured by in vivo neutron activation using the method described by Oxby et al.“‘The patient was irradiated in the supine position for 40 set on the right side and for 40 set on the left side with a sealed tube neutron generator producing 14 MeV neutrons; the total dose of radiation given for one examination was 50 mrem. The patient was immediately transferred to a whole body counter, and the gamma radiation from the induced radioactivity was measured. The complex spectrum thus obtained was analyzed by a computer program to calculate the amount of nitrogen in the body, and the value was corrected for interfering radioactivity from phosphorus, chlorine, and oxygen and for differences in body by multiplying builds. Total body protein was calculated the total body nitrogen by 6.25. In the course of measuring total body nitrogen, total body potassium was also measured by measuring the radioactivity of the naturally occurring radioisotope potassium-40.“’ For these studies, total body water was measured di-

DIET AND

Statistical

whether different

Analysis

Paired Student’s t-test was used to determine the changes within each group were significantly from zero.

Feeding

Regimes

A proprietary preparation (Flexical, Mead Johnson Laboratories) was used to feed the patients in the ED group. This provided 9% of its total energy values as protein, 61% as carbohydrate, and 30% as fat (i.e., 67% of nonprotein energy as carbohydrate and 33% as fat), and when reconstituted to “normal” strength the diet provided one kilocalorie (4.2 kilojoules) per milliliter. The diet was given by continuous infusion throughout the day and night via a fine silicone rubber nasogastric tube (internal diameter: 1.5 mm) in 10 patients, and via a fine jejunostomy plastic catheter (internal diameter: 1.00 mm) inserted at the time of surgery in 4 patients. A McGaw N7922 volumetric infusion pump (Imed Corporation, San Diego, Calif.) was used to deliver the diet in 6 patients, and gravity feeding was used in the remaining 6 patients. To enable the patients to adapt to the diet a quarter or half strength preparation of the feed was used initially at a rate of 2000 ml/24 hr, and the concentration was gradually increased to full strength over the next 2-4 days. The rate of infusion was subsequently increased where indicated to 3000-3500 ml/24 hr, depending on the size of the patient. Patients receiving intravenous nutrition were given a solution of 4.3% crystalline amino acids (Freamine II, McGaw) and 25% glucose through a central venous cath-

654

Table

YEUNG

1.

Patient

ET AL.

GASTROENTEROLOGY

Data--Indications

for Feeding

in the Two Groups Elemental

Preoperative Early

1 Pyloric stenosis 3 Abdomino-perineal 1 Proctocolectomy

postoperative

I. V. Nutrition

diet

excision

of rectum

Enterocutaneous fistula Short gut syndrome Colitis Pancreatitis

2 Large bowel 3 2 0

Other

I Failure to thrive after vagotomy 1 Depressed immunologic state with liver abscess

eter. To each liter of the solution was added 40-50 mmol of sodium, 40-60 mmol of potassium, 20 mmol of phosphate (as acid salt), and 4 mmol of magnesium. In addition, one ampul of multivitamin preparation (Multibionta, Merck & Co., Inc., Rahway, N.J.) was administered in one bottle of the nutrient solution daily, as well as appropriate intramuscular doses of vitamin K, vitamin B,,, and folate

weekly. There was no major complication related to the feeding regime in either group of patients. A few of the patients developed nausea and vomiting in the ED group, but these symptoms settled on reducing the strength of the elemental diet or the rate of infusion for a short period of time. Two patients in the IVN group developed a pyrexia 1 wk after feeding was started and had their subclavian catheter changed, though the pyrexia was thought to be due to

abdominal

sepsis rather than catheter sepsis.

Results Tables 1 and 2 show that though there was no attempt made at matching the patients, the two groups were fairly comparable in most respects. The number of patients in each of the broad categories similar in both laid down in Table 1 was roughly groups. Table 2 shows that the groups were also similar in terms of sex ratio, age, body weight, percent weight loss (calculated from recalled weight when the patients were well), and mean duration of feeding. However, the IVN patients received more calories per kg body weight per day (42.4 kcal/kg body wt/day) than the ED patients (36.3 kcal/kg body wt/ Table

2.

Patient

Data

in the Two Groups

Age (yr) Body weight (kg) Weight loss (a,) Duration of feeding (days) Energy intake (nonprotein) (kcol/kg Nitrogen intake (g/kg body wt/doy)

body wt/doy)

Vol. 77, No. 4, Part 1

1 3 1 1 1 2 2 2 1

Gastro-jejune-colic fistula Abdomino-perineal excision Total gastrectomy Reverse small bowel loop Small bowel

Failure

to thrive

after subtotal

of rectum

colectomy

day), and nearly twice as much nitrogen per kg body weight as the ED group (0.14 g/kg body wt/day, compared with 0.26 g/kg body wt/day). Table 3 shows the mean changes in body weight, fat, protein, and water for the two groups of patients, and Figure 1 shows the mean changes with +l standard error of the mean. It can be seen that in the group of patients receiving the ED, there was no significant change in any of the four components measured, i.e., body weight, fat, protein, and water. In the IVN group of patients, however, there was a mean gain of 3.2 kg in body weight; there was a smaller, but significant gain (P < 0.01) in body fat (0.7 kg); there was no significant change in body protein, but there was a mean gain of 2.4 kg in body water. The gain in body weight in the IVN patients was too large to be attributed to a gain in body fat alone, or a gain in body protein. It must therefore be due to a gain in body water. The individual changes in body weight and body water were analyzed in this group of patients, and a significant relationship was found to exist between the changes in these two components of body composition (r = 0.533, P < 0.05). The mean changes in total body potassium were +60 f 47 mmol (mean + SEM) in the ED patients and +93 + 43 mmol (mean + SEM) in the IVN patients. These changes were, however, not significantly different from zero (paired Student’s t-test). Table 4 shows the plasma albumin, prealbumin,

Studied Elemental diet (mean + SD)

I. V. Nutrition (mean + SD)

3 males : 11 females 57.0 -c 13.7 55.9 f 11.7 14.3 f. 10.0 14.9 f 2.1 36.3 + 10.8 0.14 l?r 0.04

4 males : 10 females 57.0 + 6.9 57.1 -c 10.7 12.2 f 11.8 13.9 f 1.6 42.4 f 9.6 0.26 f 0.05

0ctotx:r

ELEMENTAL

1979

Table 3.

in Two

Changes in Body Composition Body Weight ED

DIET

BODY

(kg)

Protein

Fat (kg)

IVN

ED

.______..

IVN

ED

change

0.2

3.2

0

0.7

0

0.6

0.A

0.3

0.2

0.3

Changcl

(kg)

Water

IVN -0.1 0.4

2.4 0.8

0.3

7.7

5.6

0

4.3

0

3.76

0.14

4.09

0.10

0.30

0.24

I’

NS

NS

NS

NS

ED = elumontal

diet: IVN = I.V. nutrition:

Discussion There was no attempt made in this present study to match or randomize the patients as this would have been very difficult, if not impossible, in practice. For this reason it may be argued that the two groups of patients were not ideally matched, and consequently there may be limitations in the comparisons made. Nevertheless the data presented in Tables 1 and 2 show that the similarities in the two groups in terms of sex ratio, body weight, degree of weight loss, and the broad diagnostic categories (Table 1) were close enough to enable certain observations to be made. The study showed that the administration of the

?? Elemental

I.V.

Diet

Nutrition

; I

1 A WEIGHT

Figure

A FAT

2.64 to.02

and NS = not significant

and transferrin levels in the two groups of patients before and after 2 wk of nutrition. The levels of these three plasma proteins were very similar in both groups at the start of feeding and did not change significantly with feeding. The plasma albumin and prealbumin levels at the start of feeding were at the lower end of our range for normal controls, and the transferrin levels were just below the normal range. Our range for healthy individuals were as follows: albumin 37-49 g/liter, prealbumin 18-29 mg/dl, transferrin 221-328 mg/dl (mean f 2 standard deviations).“’

*4


IVN

0.7

0.36

NS

(kg)

0.1

0.4

to.01

-1.2

ED

I-test (paircd)

(‘x,)

655

COMPOSITION

Groups Over the &Week Period

SE of changv

Mean

AND

A PROTEIN

A WATER

1. Changes in body weight, fat, protein, and water in the z groups of patients, expressed as mean + 1 SEM (kg).

elemental diet was effective in maintaining body composition in the group of 14 ill surgical patients examined and that retention of excessive water did not occur. Those patients who were fed intravenously showed an average weight gain of 5.6% over the 2-wk period, but this reflected a gain in body water rather than a gain in lean tissue mass. This finding confirms that of a previous study in which it was who shown that the weight gained by 16 patients were given a similar regime of intravenous nutrition was also due to water retention.” Two factors may help to explain this water retention in the intravenously fed patients. It has been recognized for many years that starvation or carbohydrate deprivation causes sodium and water depletion’*~‘” and that refeeding results in sodium and water retention.15.” One possibility was that our intravenously fed patients were sicker and were in a relatively more “starved” condition at the beginning than those receiving the elemental diet. If this was so, it would be reasonable to expect that the degree of sodium and water retention when feeding was instituted would be greater in the former group of patients. However, from the various indices of nutritional status assessed, there seemed to be no evidence to support the contention that one group of patients was more starved than the other. The percentage weight loss (Table 2) and the plasma protein levels (Table 4) were similar in both groups at the start of feeding. The alternative explanation lies in the difference in the calorie source of the two regimes used. It has previously been shown that the antinaturesis that occurs on refeeding the starved subject was due exclusively to carbohydrate and was not induced by fat or protein fed in isocaloric amounts.” The elemental diet used in our study provided 36.3 kilocalories/kg body wt/day of which 33% was in the form of fat. It is therefore reasonable to assume that this diet would have less antinaturetic effect than the intravenous regime which not only provided more calories (42.4 kilocalories/kg body wt/day) but in which glucose was the sole calorie source apart from protein. The precise mechanism of the sodium and water

656

Table

YEUNG

4.

Plasma

ET AL.

Protein

GASTROENTEROLOGY

Levels

Before

and After 2 Weeks of Feeding Elemental

P Plasma prcalbumin mean +- SEM) P Plasma transferrin Mean -+ SEM P

39.7 f

1.6

in the Two Groups

diet

I.V. Nutrition After

Before

After

38.9 + 0.7

39.5 -t 1.4

36.5 f 0.6

23.0 f 1.8

18.7 f 2.2

215 f 14

200 f 16

Before Plasma albumin (g/liter, mean -t SEM)

NS

(mg/dl,

19.3 * 2.5

NS

NS (mg/dl,

214 f 16

Vol. 77, No. 4, Part 1

22.5 +- 1.8

NS

NS

retention with carbohydrate refeeding is not completely understood. A decrease in ketone body production is thought to be related to the retention of sodium, but the relationship is by no means a simple one.” Hormonal factors have also been implicated. It has been shown for instance, that plasma immunoreactive insulin levels are higher in patients who are fed intravenously with a predominantly glucose system than in patients fed with a predominantly lipid system.‘” In the presence of a raised plasma insulin level, glucose is thought to be capable of increasing the reabsorption of sodium in the renal tubules.‘” It would be interesting to know how the excess water retained in the intravenously fed patients is distributed between the intracellular and extracellular spaces. In a previous study, it was shown that the intracellular potassium concentration of muscle biopsies taken before and at the end of a Zwk course of intravenous nutrition using a similar regime remained unchanged.*l Our present study has shown that total body potassium has not changed significantly in either group of patients. It can therefore be concluded that the intracellular water volume has remained unchanged, and that the increase in total body water in the intravenously fed patients must therefore be due largely to an increase in extracellular water. Since this is so and since the mean plasma sodium level before feeding (134.7 f 1.5 mmol/liter, mean + SEM) in the intravenously fed patients was not significantly different from that after feeding (136.7 f 1.2 mmol/liter), it can be concluded also that the water retention in these patients was accompanied by sodium retention. It is rather disappointing to find that neither of the two groups of patients gained significant protein and the plasma proteins were not significantly increased. It is possible that the protein gain was of a magnitude not detectable by the method of measurement used or that the calorie and nitrogen intake was inadequate for the degree of surgical trauma or stress in these patients. There is evidence from studies of body composition in severely ill surgical patients re-

226 rt 10

NS

ceiving total parenteral nutrition that the minimum caloric requirement to maintain body composition is around 40-50 kcal/kg body wt/day, and a gain in lean tissue mass is unlikely to occur unless this value is exceeded.” Our patients fed with the elemental diet received only 36.3 kcal/kg body wt/day which would be inadequate by the above criterion. The intravenously fed patients, however, received a higher caloric load (42.4 kcal/kg body wt/day), and the fact that this group gained body fat would suggest that the energy intake was adequate but there was an inability to synthesize protein. In conclusion, the technique of administration of an elemental diet by continuous infusion is comparable to central venous nutrition in maintaining body protein in severely ill patients, but has the advantage of being cheaper, safer, and easier to manage. The use of such a diet, containing a substantial proportion of its energy source as fat, is not accompanied by the problem of extracellular water retention which is seen in our intravenously fed patients who were given an “all glucose” regime. If it is properly administered it is a good alternative to intravenous alimentation in patients requiring nutritional support, provided a sufficient length of functioning small bowel is available for absorption of the diet.

References 1. Dudrick SJ, Wilmore DW. Vars HM, Rhoads JE: Long term total parenteral nutrition with growth, development and positive nitrogen balance. Surgery 64:134, 1968 2. Collins JP, Oxby CB, Hill GL: Intravenous amino acids and intravenous hyperalimentation as protein-sparing therapy after major surgery. A controlled clinical trial. Lancet 1:788, 1978 3. Hill GL, McCarthy ID, Collins JP, Smith AH: A new method for the rapid measurement of body composition in critically ill surgical patients. Br J Surg 65732, 1978 4. Stephens RV, Randall HT: Use of a concentrated, balanced, liquid elemental diet for nutritional management of catabolic states. Ann Surg 170:642,1969 5. Russell RI: Progress report: Elemental diets. Gut 16:68,1975 6. Page CP, Ryan JA, Haff RC: Continual catheter administration of an elemental diet. Surg Gyn Obstet 142:184, 1976 7. Bethel RA, Jansen RD. Heymsfield SB, Ansley JD, Rudman D:

October

ELEMENTAL

1979

Nasogastric hyperalimentation eter: an alternative to central

through a polyethylene venous hyperalimentation.

cathAm J

Clin Nutr 31:698, 1978 8. Durnin JVGA, Wormersley J: Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br J Nutr 3277.1974 9. Siri WE: Univ Calif Radiat Lab Pub1 No. 3349, 1956 10. Oxby CB, Appleby DB, Brooks K, Burkinshaw L, Krupowicz DW. McCarthy ID, Oldroyd B, Ellis E, Collins JP, Hill GL: A technique for measuring total-body nitrogen in clinical investigations using the “N(n,Zn) 13N reaction. Int J Appl Radiat Isot 29:205, 1978 11. Bartholomew RL, Delaney A: Spectrophotometric studies and analytical application of the protein error of some pH indicators. Proc Aust Assoc Clin Biochem l&l,1964 12. Mancini G. Carbonara AO. Heremans JF: Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 2235.1965 13. Young GA, Hill GL: Assessment of protein-calorie malnutrition in surgical patients from plasma proteins and anthropometric measurements. Am J Clin Nutr 31:429, 1978 14. Benedict FG: A study of prolonged fasting. Carnegie Inst. Washington, Publ. No. 203,1915

DIET

AND

BODY

COMPOSITION

ti57

15. Gamble JL, Ross GS, Tisdall FF: The metabolism of fixed base during fasting. J Biol Chem 57:633,1923 16. Bloom WL, Mitchell W: Salt excretion of fasting patients. Arch Intern Med 106:321,1966 17. Bloom WL: Inhibition of salt excretion Intern Med 109:80,1962

by carbohydrate.

Arch

18. Veverbrants E, Arky RA: Effects of fasting and refeeding I. Studies on sodium, potassium, and water excretion on a constant electrolyte and fluid intake. J Clin Endocrinol Metab 29:55, 1969 19. Kolanowski J: On the mechanisms of carbohydrate-induced sodium lisme 3:131, 1977

of fasting natriuresis and retention. Diabete Metabo-

20. Jeejecbhoy KN, Anderson GH, Nakhooda AF, Greenberg CR, Sanderson I, Marliss EB: Metabolic studies in total parenteral nutrition with lipid in man. Comparison with glucose. J Clin Invest 57:125,1976 21. King RFJG, Collins JP, Morgan DB, Hill GL: Muscle chemistry of critically ill surgical patients and the effects of a course of intravenous nutrition. Br J Surg 65:495, 1978 22. Spanier AH, Shizgal HM: Caloric requirements of the critically ill patient receiving intravenous hyperalimentation. Am J Surg 133:99, 1977